ACVPred: Enhanced Prediction of Anti-Coronavirus Peptides by data augmentation combined with Transfer Learning

Datasets:

• ACVPs.csv: 154 ACVPs collected from ACovPepDB, PreAntiCoV and iACVP.
• AVPs.csv: 1248 AVPs collected from PreAntiCoV and iACVP.
• nonAVPs.csv: 6296 non-AVPs collected from PreAntiCoV and iACVP.
• independent_test.fasta: contains 16 ACVPs and 16 non-AVPs

Codes and model:

• augmentation.py: augmentation methods utilized in ACVPred
• model.py: prediction model structure and details in ACVPred
• predict.py: run this script to use ACVPred to apply prediction
• model.pth: model parameters

Usage:

run python predict.py -i/--input [input_file] -o/--output [output_file] in command

• input_file: route to your file containing peptides to be predicted in FASTA format.
• output_file: route to your result file containing prediction results in CSV format.
• An example dataset is given in example.fasta